ONS changes RPI to better reflect housing costs

The Office for National Statistics (ONS) is changing the way in which it calculates Retail Price Inflation (RPI) in an attempt to reflect the cost of housing more realistically.

When applied retrospectively the move drastically alters the inflationary landscape to date.

According to the ONS, using the old mortgage figures, RPI deflation last April stood at 1.2%. Using the new measure, it would have stood at a positive 0.2%.

RPI is used to determine a huge range of pricing, including public sector pay, index-linked gilts and the State Pension.

It measures a basket of goods including mortgage interest, council tax and certain other housing costs to give a measure of the cost of living.

In the past, the mortgage interest calculation was based on all borrowers paying their lender's standard variable rate (SVR).

The new way of measuring mortgage interest costs relies on data from the Bank of England that covers a weighted average of fixed, tracker, discount, and other mortgages, which it says make up about 90% of the mortgage universe.

The ONS said the change was being made "because the new rate is more representative of the mortgage rates available".

Ray Boulger, senior technical manager at mortgage brokerage John Charcol, says: "It makes sense to seek a better average measure for RPI, as this figure has been unreliable for some time. The problem is that this attempt appears cackhanded to say the least.

"The timing is utterly bizarre, given that there are actually more borrowers paying their lender's SVR than at any time in the past, and the proportion is only likely to grow over the next couple of years. So choosing now to change the measure in this way and using 10% as the assumed benchmark proportion of mortgage borrowers on SVR is peculiar. I suspect some of the larger lenders now have around 50% of their borrowers paying SVR.

"I am not even convinced that the Bank of England has the correct information from which to calculate the proposed weighted average.

"The worry is that this change will simply bring about a move from one unreliable RPI statistic to another."